Update README.md
Browse files
README.md
CHANGED
@@ -5,15 +5,50 @@ license: mit
|
|
5 |
## Usage
|
6 |
|
7 |
```python
|
8 |
-
|
9 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
10 |
model.eval()
|
11 |
|
12 |
-
|
13 |
-
|
|
|
|
|
|
|
|
|
|
|
14 |
|
|
|
|
|
|
|
15 |
with torch.no_grad():
|
16 |
-
out = model(
|
17 |
|
18 |
probabilities = F.softmax(out[0], dim=0)
|
|
|
|
|
19 |
```
|
|
|
5 |
## Usage
|
6 |
|
7 |
```python
|
8 |
+
import torch
|
9 |
+
from torch import nn
|
10 |
+
import torchvision.transforms as transforms
|
11 |
+
import torch.nn.functional as F
|
12 |
+
from pathlib import Path
|
13 |
+
|
14 |
+
LABELS = Path("classes.txt").read_text().splitlines()
|
15 |
+
num_classes = len(LABELS)
|
16 |
+
|
17 |
+
model = nn.Sequential(
|
18 |
+
nn.Conv2d(1, 64, 3, padding="same"),
|
19 |
+
nn.ReLU(),
|
20 |
+
nn.MaxPool2d(2),
|
21 |
+
nn.Conv2d(64, 128, 3, padding="same"),
|
22 |
+
nn.ReLU(),
|
23 |
+
nn.MaxPool2d(2),
|
24 |
+
nn.Conv2d(128, 256, 3, padding="same"),
|
25 |
+
nn.ReLU(),
|
26 |
+
nn.MaxPool2d(2),
|
27 |
+
nn.Flatten(),
|
28 |
+
nn.Linear(2304, 512),
|
29 |
+
nn.ReLU(),
|
30 |
+
nn.Linear(512, num_classes),
|
31 |
+
)
|
32 |
+
|
33 |
+
state_dict = torch.load("model.pth", map_location="cpu")
|
34 |
+
model.load_state_dict(state_dict)
|
35 |
model.eval()
|
36 |
|
37 |
+
transform = transforms.Compose(
|
38 |
+
[
|
39 |
+
transforms.Resize((28, 28)),
|
40 |
+
transforms.ToTensor(),
|
41 |
+
transforms.Normalize((0.5,), (0.5,)),
|
42 |
+
]
|
43 |
+
)
|
44 |
|
45 |
+
def predict(image):
|
46 |
+
image = image['composite']
|
47 |
+
tensor = transform(image).unsqueeze(0)
|
48 |
with torch.no_grad():
|
49 |
+
out = model(tensor)
|
50 |
|
51 |
probabilities = F.softmax(out[0], dim=0)
|
52 |
+
values, indices = torch.topk(probabilities, 5)
|
53 |
+
print(values, indices)
|
54 |
```
|